The study's goal was to see whether the environmental repercussions of massive infrastructure projects might be tracked using remote sensing data (satellite photos) and the training of convolutional neural network models. The paper is concerned with environmental monitoring throughout the development and operation of the belt of the Chinese Railway's vast transport infrastructure (China Railway Express), which links Asia and Europe.
"Through this scientific research, we clearly demonstrated the high applicability of deep learning methods for monitoring environmental conditions, where we concentrated on recognizing the impact of significant infrastructure projects, however the provided model can have a much broader use for detecting and monitoring diverse environmental changes. Furthermore, the utilization of remote sensing, i.e. satellite photos from publicly available and free Sentinel missions, demonstrates a substantial advantage over conventional methods used to monitor the status of the environment.
This is the first step in our future research, as we have looked at the possibilities and assessed the potential of remote sensing technologies in combination with neural networks. We are able to identify subsequent research lines, which will involve the publishing of new scientific works, and this will be important not only for me as a junior researcher, but will also contribute to the ongoing growth of our institute," says Marko Pavlović, a junior researcher at the Institute of Artificial Intelligence of Serbia.
The published paper is an important scientific addition in the field of remote sensing, and the research findings demonstrate that there is a large potential for effective monitoring and analysis of environmental changes using deep learning methods.
"Remote Sensing" is a peer-reviewed, open access journal about the science and application of remote sensing technology. It is part of the "MDPI" publication group, which has published research by over 330,000 individual authors and is supported by over 115,000 academic specialists worldwide, and the scientific paper "Monitoring the Impact of Large Transport Infrastructure on Land Use and Environment Using Deep Learning and Satellite Imagery" is available here.